Gram-Elites: N-Gram Based Quality-Diversity Search

Biemer, Colan F.; Hervella, Alejandro; Cooper, Seth

doi:10.1145/3472538.3472599

Cited by 6 publications

(5 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Gram-Elites: N-Gram Based Quality Diversity Search Gram-Elites (Biemer, Hervella, and Cooper 2021) builds on top of MAP-Elites by modifying genetic operators to be based on n-grams, see the code 1 or paper for more information. The result is that every level generated is guaranteed to be well-formed, meaning there are no broken in-game structures, such as malformed pipes in Mario.…”

Section: Work To Datementioning

confidence: 99%

See 1 more Smart Citation

Dynamic Difficulty Adjustment via Procedural Level Generation Guided by a Markov Decision Process for Platformers and Roguelikes

Biemer

2023

AIIDE

View full text Add to dashboard Cite

Procedural level generation can create unseen levels and improve the replayability of games, but there are requirements for a generated level. First, a level must be completable. Second, a level must look and feel like a level that would exist in the game, meaning a random combination of tiles that happens to be completable is not enough. On top of these two requirements, though, is the player experience. If a level is too hard, the player will be frustrated. If too easy, they will be bored. Neither outcome is desirable. A procedural level generation system has to account for the player's skill and generate levels at the correct difficulty. I address this issue by showing how a Markov Decision Process can be used as a director to assemble levels tailored to a player's skill level, but I've only demonstrated that my approach works with surrogate agents. For my thesis, I plan to build on my past work by creating a full roguelike and platformer and running two player studies to validate my approach.

show abstract

Section: Work To Datementioning

confidence: 99%

“…My work started by generating level segments for platformers and roguelikes (Biemer, Hervella, and Cooper 2021) using a modified version of MAP-Elites (Mouret and Clune 2015). A full level is assembled by concatenating level segments together.…”

Section: Introductionmentioning

confidence: 99%

Dynamic Difficulty Adjustment via Procedural Level Generation Guided by a Markov Decision Process for Platformers and Roguelikes

Biemer

2023

AIIDE

View full text Add to dashboard Cite

show abstract

“…Gram-Elites [9] is an extension to MAP-Elites [26] that uses n-grams for population generation and the genetic operators mutation and crossover. These operators use a concept called connection, which uses a breadth-first search through the ngram to ensure that post-modification the new segment will be generable.…”

Section: A Gram-elitesmentioning

confidence: 99%

“…We test with three games: Mario, a horizontal platformer, Kid Icarus, a vertical platformer, and DungeonGrams, a topdown roguelike. 2 For all three games, we link level segments generated by Gram-Elites [9]. 3 We chose Mario as a baseline because much research in procedural content generation via machine learning for games uses it.…”

Section: Introductionmentioning

confidence: 99%

On Linking Level Segments

Biemer¹,

Cooper²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

An increasingly common area of study in procedural content generation is the creation of level segments: short pieces that can be used to form larger levels. Previous work has used basic concatenation to form these larger levels. However, even if the segments themselves are completable and well-formed, concatenation can fail to produce levels that are completable and can cause broken in-game structures (e.g. malformed pipes in Mario). We show this with three tile-based games: a side-scrolling platformer, a vertical platformer, and a top-down roguelike. Additionally, we present a Markov chain and a tree search algorithm that finds a link between two level segments, which uses filters to ensure completability and unbroken in-game structures in the linked segments. We further show that these links work well for multi-segment levels. We find that this method reliably finds links between segments and is customizable to meet a designer's needs.

show abstract

“…A challenge for training on dungeon-based games is the lack of data in the VGLC where the only such game with text-based level data is The Legend of Zelda (other Zelda games are in a graph format). Thus, to be able to blend multiple such games, we also used levels from DungeonGrams (DGG), a roguelike dungeon crawler developed for prior research[8] as well as the previously used Metroid levels.While Metroid is a platformer, its sprawling, interconnected world can be seen as a dungeon when viewed from a top-down perspective. As before, we trained on 15x16 segments from each game.…”

mentioning

confidence: 99%

Learning latent representations for controllable combinational creativity and game design

Sarkar

View full text Add to dashboard Cite

Gram-Elites: N-Gram Based Quality-Diversity Search

Cited by 6 publications

References 10 publications

Dynamic Difficulty Adjustment via Procedural Level Generation Guided by a Markov Decision Process for Platformers and Roguelikes

Dynamic Difficulty Adjustment via Procedural Level Generation Guided by a Markov Decision Process for Platformers and Roguelikes

On Linking Level Segments

Learning latent representations for controllable combinational creativity and game design

Contact Info

Product

Resources

About